Hello and welcome! It's Daniel here, the author of LangChain for JavaScript Developers.
Today, I’m excited to share my interview with Ishan Anand, whom I discovered while watching recordings from this year's AI Engineering World's Fair.
Ishan shared how he implemented a working GPT-2 model directly into Excel. Yes, you heard that right Excel. Ok, a very big Excel file, but still... just a pure spreadsheet.
I was fascinated by this because to implement a model like that, you need to really understand how it works down to the smallest details, not just have a general overview.
I believe that once you can do this, you become a better engineer. It deepens your understanding of what LLMs can do, helps you interpret research papers, and allows you to be the rational voice when others attribute magical powers to AI.
So, enjoy this interview with Ishan! He’s a great teacher and has a lot of valuable information to share. Let’s dive in!
💻 Mentioned Links
Below are the links mentioned during our talk:
- Spreadsheets Are All You Need
- Ishan course on Maven - get 20% off using the JSCRAFT promo code
- Ishan's Twitter account
- Decoding the Decoder LLM without de code
- Attention Is All You Need research paper
Recommend learning resources:
📖 Neural Networks from Scratch - Presale
I'm writing a book about the timeless foundational concepts of neural networks for JavaScript developers. Go from if-else to weights and biases by building tiny AI models from scratch!
📖 Neural Networks from Scratch - Presale
I'm writing a book about the timeless foundational concepts of neural networks for JavaScript developers. Go from if-else to weights and biases by building tiny AI models from scratch!